Oleksandr Radomskyi

Hi @MatthewBM , @mk123qwe do you know how to run ffn on a dataset of about 5TB?

Hi @subeeshvasu, if the question is still actual for you: recently I managed to test both consensus and resegmentation. in the heart of consensus is boolean AND comparison between two...

Hi @Aliengmx add import line below to the beginning of file bounding_box.py from typing import Optional

Solution: Remove or replace all non-ascii characters from train data. On Windows OS: 1. Install Notepad++ from here: https://notepad-plus-plus.org/download/v7.3.html 2. Open chat.in in Notepad++ and do follow: 3. Ctrl+F 4....

@dsblr Solution: 1. Copypaste file "tf_seq2seq_chatbot/data/train/movie_lines_selected.txt" 2. Rename it into "chat.in" 3. In ../tf_seq2seq_chatbot/configs/config.py Specify absolute path to "chat.in" using Windows style with special characters enclosing (dash '\\' is special...

@dsblr you can try manually search and clean file, but I prefer cleaning with regex. also, you can try Encoding -> Convert to ANSI and then save character 0xad at...

still training... single gtx950, 6 days, global step is above 950 000.. interesting to see result..